USBc rail filtering with ferrite bead for voltage regulators

Question

I am creating a mixed signal PCB which has two separate supplies: A buck converter for the digital section and an LDO Regulator for the analog one.

Each regulator steps down 5 V from a USB-C supply to 3.3 V. 
The digital section draws up to 3A, while the analog section draws up to 500 mA.

For the digital side, I am using a TPSM82864AA0SRDJR Buck Converter (lower regulator on the picture, 2.4 MHz switching frequency) and for the analog section, an AMS1117-3.3V, both powered by the VBUS supply line (USB-C, see schematic on the top left).

I have placed a Pi filter consisting of two 1 µF capacitors and a ferrite bead (capable of handling the needed current) in order to suppress the high-frequency noise from the (usually very noisy) USB-C supply rail before reaching my voltage regulators.

I want to ask if this Pi filter arrangement is a good idea considered my regulator setup, or if a different approach would be more appropriate.

Thank you very much in advance.
Greetings,
 Leandro.

Something like this is far more readable. It's not perfect - lines could be thicker and general visibility could be improved, but it gives an idea.

Answer 1

Summary:

The ferrite bead might not be a bad idea, but neither is it likely a good idea, or strictly necessary. There are other questions you may wish to ask about your circuit, but which are outside of the scope of the exact question asked here.

---

Ferrite Bead

I would not worry about the ferrite bead. Such a small value has little filtering effect, and Murata does not provide characteristic data at all on ferrite beads under DC bias. Typically a ferrite bead saturates (impedance at relevant frequencies drops say 30% or more) in the 50 to 200mA range; higher for larger chips and lower impedances, and perhaps the BLM18SG330SN1 saturates as high as an ampere, but you want over 3 so it will still have less contribution than you were initially hoping.

Where you know filtering is necessary, choose an inductor instead. This is an inductive component with lower losses and rated saturation current. For DC bias below the saturation point, ample impedance is available. The loss of the ferrite bead may actually be desirable, but when the saturation is not available, we must use an inductor proper (the loss can be substituted by using a parallel resistor, or lossy capacitors -- typically electrolytics).

USB Standards, Ratings

Beware that you must honor the bulk capacitance and inrush current limits of the relevant standard. As you mention 5V 3A, I'm guessing this is negotiated via USB-PD, Type-C. In particular, cSnkBulk and cSnkBulkPd limits apply. See: USB Power Delivery | USB-IF section 7.2

You may want transient protection, in accordance with §7.2.9.2. Compare relevant articles such as:
Correct Placement of Series Ferrite Beads to Avoid DC Disconnect During Power Cycling

I don't see reference to current ripple requirements, and it is possible they do not specify emissions; or perhaps more to the point, that functional limits define it: excessive AC current flow in the cable might cause excessive temperature rise; insufficient local bypass or damping causes instability of the source and/or sink; etc.